Increasing utilization of limb sparing procedures for malignant or aggressive benign bone and soft tissue tumors frequently results in large skeletal defects in bone. The most common reconstructions involving the lower extremity include the use of osteoarticular allograft about the knee and a combination of allograft with metallic implant for proximal femur replacement. Several problems are associated with these procedures making their clinical outcome unpredictable. The underlying hypothesis of this study is that some of these problems are surgical and biomechanical in nature and can be resolved by better size matching of the graft to the host, improved implant design, and the use of an effective graft/prosthesis fixation method. The main goal of this extension project is to improve both the clinical and functional results of patients undergoing reconstruction of large bony defects around the knee and proximal femur using osteoarticular allograft and allograft/prosthetic composites. To accomplish this goal, we propose: 1) to develop an allograft size screening and shape matching program for selecting ideal allografts to fit the host bone defects, allowing near normal joint kinematics and contact pressure distribution; 2) to design and test a long-stem femoral prosthesis system to be used specifically with an allograft for proximal femur replacement; and 3) to follow up the clinical, radiographic, and functional results of patients with massive allograft transplantation. Osteoarticular bone dimension and curvature of a large population and bone bank stock will be measured using both radiographs and spatially reconstructed CT scan to form a macro-size screening scheme for allograft selection. 3-D joint kinematic analysis and joint surface contact pressure distribution will be utilized to formulate an additional microsize matching scheme. These analysis results will be used to establish a practical graft size matching index for clinical application. Using theoretical modeling and experimental testing, the effect of surgical placement of osteoarticular allograft on joint motion and pressure distribution will also be studied. An improved allograft and hip endoprosthesis fixation method for the proximal femur replacement will be tested using in vitro and in vivo canine models. The data obtained from the canine model will be used to design the human implant and validated by applying the implant and the optimal fixation technique to cadaveric femora. Radiographic analysis following the new evaluation criteria recommended by the American Musculoskeletal Tumor Society will be performed. Both objective and subjective functional assessments will be adopted for patient follow-up. The results of these studies are expected to improve the clinical outcome of massive allograft transplantation procedures in orthopedic surgery.